0% found this document useful (0 votes)
18 views7 pages

Data Science - Machine Learning - Terminology

Uploaded by

Majety S Lskshmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
18 views7 pages

Data Science - Machine Learning - Terminology

Uploaded by

Majety S Lskshmi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Data Science – Machine Learning - Terminology

2. Data Science – Machine Learning – Terminology

Contents
1. Machine learning ................................................................................................................................ 2
2. Model .................................................................................................................................................. 2
3. Train .................................................................................................................................................... 3
4. How to train? ...................................................................................................................................... 3
5. Model learns ....................................................................................................................................... 3
6. Pattern ................................................................................................................................................ 3
7. Before trained .................................................................................................................................... 3
8. After trained ....................................................................................................................................... 3
9. Test the model.................................................................................................................................... 3
10. Model deployment ........................................................................................................................... 4
11. Towards target value or classifier .................................................................................................... 4
12. General example .............................................................................................................................. 4
13. Smart application ............................................................................................................................. 4
14. Normal application ........................................................................................................................... 5
15. Smart application ............................................................................................................................. 6
16. Use case ............................................................................................................................................ 6
17. Accuracy testing ............................................................................................................................... 7

1|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

2. Data Science – Machine Learning – Terminology

1. Machine learning

 Machine learning is a technique which enables computers to learn


automatically from past data.
 Machine Learning is an approach to train the models.
 Machine learning is helpful to predict the future values.

2. Model

 In simple terms, a model is a representation of reality.


 A model is created to do a specific purpose

Example
 General example, a world map is a model for this physical world or
universe.
o The creator of world map have done deep analysis, understands,
visualized with simple representation.

In terms of Machine learning,

 A model can be a,

o A piece of code or program


o A mathematical formula
o A program + mathematical formula

2|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

3. Train

 We need to train the model

4. How to train?

 We need to use data to train the model

5. Model learns

 Model will learn pattern from data.

6. Pattern

 Patterns means a simple knowledge gained during training.


 So, model learnt patterns from data during training

7. Before trained

 Before training model, model doesn’t know about the patterns.

8. After trained

 After training model, model knows about the patterns.

9. Test the model

 Once after model got trained then that model has to be tested to check
the accuracy.
 If accuracy is good than the model works well

3|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

10. Model deployment

 Finally we need to deploy knowledge acquired by the model


 These models also called as model fit

11. Towards target value or classifier

 Models learnt patterns from data towards target value.


 So, let’s try to understand what it means.

12. General example

 Assuming that one of my friend is introducing his friend as


o His name is Prasad working in X Company and having 3 years of
experience.
 When I heard about his experience I can guess/predict his salary (target
value).
 So, out brain is already trained to guess some attributes from past
data(old friends)
 So, we have patterns towards to target value

13. Smart application

 Once accuracy satisfied then we need to deploy the model.


 Once model deployed then that application will become as smart
application

4|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

14. Normal application

 Assuming that we have created an application by using


Java/dotnet/Python.
 It’s just called as normal software application because this application is
just works on instruction based.
 Instruct based means,
o If x value is greater than y value, then prints x value
o If x value is less than y value, then prints y value

Make a note
 Normal application cannot think like human to take decisions
 Normal applications cannot take own decisions

5|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

15. Smart application

 A smart application is an application which thinks like a human to take


the decisions.
 Smart application is intelligence based.

16. Use case

 Whenever you send email automatically Gmail application is predicting


weather it is a spam or not spam.
 That means Gmail application is well trained and captures the patterns
about mail type like,
o Spam mail
o Promotion mail
o Social media mail & etc
 So, Gmail application knows patterns about to recognize the type of
mail.

6|Page 1.ML - Terminology


Data Science – Machine Learning - Terminology

17. Accuracy testing

 Before deploying the model, it needs to be tested in lot of dimensions.


 If model is getting good predictions then we need to deploy.

Process in machine learning

 We need to train the model.


 During training model acquires the knowledge.
 This knowledge is called as model fit
 Once after model fit produced we need to test to check the accuracy.
 If satisfied with accuracy then we need to deploy the model

Data is important

 We need to train the model with data


 During this training model will go through the data and understand
the patterns about data.
 These patterns helps to predict the result on new data

7|Page 1.ML - Terminology

You might also like